Virtual-machine technology essentially abstracts the hardware resources and interfaces of a computer system on behalf of one or multiple virtual machines, each comprising one or more application programs and an operating system. The recent emergence of cloud computing services can provide abstract interfaces to enormous collections of geographically dispersed data centers, allowing computational service providers to develop and deploy complex Internet-based services that execute on tens or hundreds of physical servers through abstract cloud-computing interfaces.
Within virtual servers as well as physical servers, virtual machines and virtual applications can be moved among multiple virtual or physical processors in order to facilitate load balancing and to co-locate compatible virtual machines and virtual applications with respect to virtual and physical processors. Similarly, virtual machines and virtual applications can be moved among the virtual servers in a virtual data center as well as among physical servers in the underlying physical hardware in which virtual data centers are constructed. Migration of virtual machines and virtual applications in virtual data centers can also be used for load balancing, fault tolerance and high availability, and for many other purposes.
However, managing the virtual servers as well as physical servers, virtual machines and virtual applications often requires on site access which can be expensive and inconvenient depending upon the location of the remote site.